Technical limitations and recommendations for streamflow predictions in data-scarce regions using GloFAS-ERA5 reanalysis data

Anadyr River case study

Anatoly Tsyplenkov & Danila Shkolny
Lomonosov Moscow State University

What does a boa constrictor that has eaten an elephant look like?

What does a boa constrictor that has eaten an elephant look like?

Anadyr River is the largest ungauged¹ Russian River

Nota bene

No streamflow measurements on the state gauging station network in the last 30 years!

¹no streamflow gages since 1996

Anadyr River is the largest ungauged¹ Russian River

What can we do?

How can we predict streamflow?

How accurate are our predictions?

¹no streamflow gages since 1996

Anadyr River is the largest ungauged¹ Russian River

What can we do?

How can we predict streamflow?

How accurate are our predictions?

We are probably late with this presentation for 10 years because IAHS PUB Decade was in 2003–2012…

¹no streamflow gages since 1996

GloFAS-ERA5 quick overview

  • Current version GloFAS-ERA5 v4.0 (June 2025 release)

  • Global hydrological models HTESSEL¹ and LISFLOOD v4.1.3

  • Daily discharge in (m³/s) at 0.05° spatial res (≈5 km)

¹Hydrology Tiled ECMWF Scheme for Surface Exchanges over Land

GloFAS-ERA5 calibration

Nearly 2,500 gauging stations used for calibration in v4.0. Global median KGE' is 0.7

We are not affiliated with the ECMWF

GloFAS-ERA5 calibration

Median KGE' for Chukotka and Kamchatka rivers — 0.9-1.0

We are not affiliated with the ECMWF

Is it really that good?

👀